Collaborative efforts of businesses have gained significant importance
in recent years. These efforts are illustrated in organizations such as
alliances. Alliances, which are « relatively enduring interfirm cooperative arrangements » (Parkhe, 1991), are often formed with an intent of
strengthening the partners’ competitive position and improving their
ability to take advantage of market opportunities. These outcomes can
be accomplished through leveraging the partners’ resources as well as
« through entrepreneurial (…) actions [geared towards] ongoing modifications to adapt to new conditions and skill requirements » (Madhok
and Tallman, 1998: 336). The latter processes relate to the organization’s dynamic capabilities that reflect the ability to seek new resource
combinations and relevant routines focusing on endeavours to facilitate and shape learning (Teece, 2003). Gudergan, Devinney and Ellis
(2002, 2003) find support for the relevance of dynamic capabilities in
alliances and show that the creativity and learning occurring in these
organizations affect their performance.
The academic literature on alliances recognizes clearly the different
functioning of different ownership, governance and organizational
structures in cooperative arrangements (e.g., Gulati and Singh, 1998).
However, there is no comprehensive understanding of the means by
which alliance organization affects the creativity and learning in
alliances. From the general management literature, we know organizational structure influences innovation processes such as learning
and creativity (Burns and Stalker 1961). Yet, the alliance literature concentrates nearly exclusively on ownership and governance structures
and their impact on managing appropriation costs and concerns when
leveraging resources in strategic alliances (e.g., Gulati and Singh,
1998) and the developmental processes of cooperative relationships
and related organizational structures (Ring and Van de Ven, 1992,
1994), but places less notice on the equally important effect of the
organizational structures that characterize an alliance on creativity and
learning occurring in the cooperative arrangement.
This paper explores how organizational settings influence the dynamic capabilities within the context of alliances. We discuss and integrate
the learning and creativity processes underlying dynamic capabilities
and relevant facets of alliance organization. Our work integrates the literatures relating to organizational structure and the learning and creativity processes in alliances. Our results suggest that centralized
structures in alliances hinder creativity and learning, and that formalization impedes learning in alliances.
To test the hypothesized effects of organizational structure on creativity and learning in alliances empirically, we employed a questionnaire
survey methodology and estimated the hypothesized relationships
using Partial Least Squares (PLS) analysis. This survey and analysis
phase was preceded by an initial set of qualitative interviews.
The setting of this research focuses on a cross-industry sample of
4,500 medium-to-large-sized organizations which have formed at least
one alliance—including both equity and non-equity alliances. To
ensure sufficient variance in the data, the sample was selected to
include organizations operating in diverse manufacturing and service
arenas. The use of this sampling procedure introduced deliberate variations in relationships under examination. The unit of analysis in this
research is the alliance team, which is formed as a result of the collaboration between two or more independent organizations. The questionnaire was directed to the alliance manager (key informant), which
ensured collection of appropriate data. The key informants were asked
to complete the questionnaire based on (only) one alliance with which
they had experience and have detailed knowledge about. This method
ensures the unit of analysis is adequately addressed and that relevant
information is collected.
The first phase of our empirical research—the set of qualitative interviews—consisted of in-depth personal interviews with experienced
alliance managers. There are three main objectives for these qualitative interviews, namely 1/to assess the relevance of the theoretical and
substantive issue under investigation, 2/to ascertain the appropriateness and feasibility of the defined sampling frame, and 3/to conduct a
preliminary test of the questionnaire for content validity and questionnaire wording.
We conducted detailed interviews with nine managers to gain an
understanding of the phenomenon of interest; the creativity and learning processes in and structural settings of the alliance team. To conduct this, we randomly identified organizations. The alliance managers
in these organizations were personally contacted over the telephone to
determine the appropriateness of their organization for the purpose of
the study and to set up an interview. The average duration of these
interviews was about 90 minutes. We also administered a preliminary
draft of our questionnaire to these alliance managers. They were
asked to evaluate the wording of the statements used and identify
other problems. Their opinions were used to refine the questionnaire.
This part of our empirical study provided initial support for the hypothesized relationships and resulted in measurement scales that had content validity.
The data collection method of personalized packages in the final survey was complemented by a reminder fax, designed to increase the
response rate. The response rate of the survey was 12.5%, which is
sufficient for testing the hypothesized theoretical relationships.
Appendix 1 provides descriptive statistics for our study.
The questionnaire consisted of reflective and formative measures.
Reflective measures were evaluated using reliability analysis (Nunnally and Bernstein, 1994) and Cronbach alpha scores, and formative
measures using Vanishing Tetrads (Bollen and Ting, 2000). The concepts and measurement scales are reported in the following paragraphs, with the individual items and scale characteristics being reported in Tables 1 and 2.
Alliance learning refers to the acquisition and development of information and skills within the alliance team. It is understood as a set containing heterogeneous, independent components. The components
that were relevant, according to the literature and the interviews undertaken with alliance managers, consisted of the following: administration or managerial techniques/practices/policies, political and legal
aspects, cultural aspects, marketing techniques, product related technologies, competitor products, customers, and production/manufacturing. Therefore, the construct was measured by an eight-item, seven-
point formative scale ranging from “People in this alliance don’t spend
any effort in learning about…” to “People in this alliance spend considerable effort in learning about…” addressing the abovementioned
Table 1 - Reflective Scale Cronbach Alphas
Cronbach alpha (for scale)
Please indicate the extent to which you agree with the following statements:
- In this alliance considerable resources are spent in developing procedures and policies for all jobs;
- People in this alliance rely on formal policies to guide decision making;
- People in this alliance must rigidly follow procedures;
- Individual decision makers at all management levels have wide latitude in making job-related choices
Please indicate the extent to which you agree with the following statements:
- Top level management makes all the important decisions and then
- In this alliance important decisions are made only at high management levels;
- This alliance is very hierarchically organised
In this alliance we…
- experiment with non-traditional methods in our decision making;
- brainstorm for “out of the square” solutions;
- have a positive attitude towards creativity;
- have novel solutions during decision making
Alliance creativity is indicative of the generation of unique ideas and
solutions within the alliance team. This construct was measured by a
four-item, seven-point reflective Likert-type scale ranging from “completely inaccurate” to “completely accurate”. These items included, “In
this alliance we experiment with non-traditional methods in our decision-making”; “… we brainstorm for ‘out-of-the-square’ solutions”; and
“… we have radically different approaches to decision-making”.
Structural formalization is defined as an inflexible system of control
governing the alliance team reflected in norms and procedures. This
construct is measured using four reflective items with a seven-point
Likert scale ranging from “strongly disagree” to “strongly agree”. They
include the following statements: “In this alliance considerable
resources are spent in developing procedures and policies for all jobs”;
“People in this alliance rely on formal policies to guide decision-making”; “People in this alliance must rigidly follow procedures”; and “Individual decision makers at all management levels have wide latitude in
making job-related choices”. All of these items address the reliance on
norms and procedures within the alliance team. For instance, “wide latitude in making job-related choices” reflects rigidity based on norms
Alliance centralization refers to the concentration of decision-making
within a small group of people within the alliance team. This construct
is measured using a reflective, four-item Likert scale ranging from
“strongly disagree” to “strongly agree”. This scale includes the following statements: “Top level management makes all the important decisions and then delegates tasks”; “In this alliance important decisions
are made only at high management levels”; and “This alliance is very
hierarchically organised”. Our preliminary interviews of alliance managers confirmed that respondents considered all statements to focus
on the decision-making concentration within the alliance team.
A range of control variables were included in our estimation to account
for their effects on creativity and learning in alliance teams (see
Appendix 2 for a detailed report on the effects on control variables).
Table 2 - Formative Scale p-values (Vanishing Tetrads)
People in this alliance don’t spend any (… considerable)
effort in learning about…
- administration or managerial techniques/practices/policies;
- political and legal aspects;
- cultural aspects;
- marketing techniques;
- product related technologies;
- competitor products;
*: d.f.: degrees of freedom
These factors included constructs such as extrinsic and intrinsic motivation, communicative interaction and diversity of alliance team members. All of these are discussed in the literature to have an influence
on creativity and learning. However, as the focus of our paper is on
whether organizational structure affects creativity and learning in
alliances, we have limited our discussion to these aspects only.
Structural equation modelling (SEM) (Jöreskog and Sörbom, 1982)
employing Partial Least Squares (PLS) analysis (Chin, 1998) was
used to estimate the hypothesised relationships. This allows making
explicit assumptions regarding constructs and theorized relationships
(e.g., Hulland, 1999)—a benefit leading to increased use of SEM in
management when accounting for multiple relationships and modelling
of unobservable variables (e.g., Bagozzi, 1980; Fornell and Cha,
1994). The complexity of the model (i.e., combination of hypothesized
effects and those of the control variables) requires the employment of
The results of our analysis provide interesting insights which we will
discuss in the following sections. Table 3 briefly summarizes the path
coefficients and significance levels for the hypotheses postulated in
Hypothesis 1 suggests that greater structural centralization in the
alliance team is associated with low levels of creativity in that team.
Supporting this hypothesis, the empirical results show that structural
centralization has a negative and significant effect on alliance creativity (path coefficient: –0.112; t-statistic: –2.359). These findings confirm
the literature that suggests that high centralization can restrict interaction and exchange and, in turn, impede creative behaviour.
Hypothesis 2 proposes that greater structural centralization in the
alliance team is associated with low levels of learning in that team. Reinforcing this hypothesis, the empirical results demonstrate that, in fact, a
negative and significant effect of structural centralization on alliance
learning can be found (path coefficient: –0.085; t-statistic: –2.969). This
result is consistent with the literature which argues that high levels of
structural centralization restrict decision-making, interaction and information exchange between members; thus, limiting learning.
Table 3 - Estimation of Effects (Partial Least Squares)
Effects on Alliance Creativity
Effects on Alliance Learning
†: One-tailed test; ****: p>0.001; ***: p>0.01
Hypothesis 3 articulates that greater structural formalization relates to
low levels of creativity in the alliance team. Our empirical results do not
provide support for this hypothesis. The results are inconclusive as the
effect of structural formalization on alliance creativity is not significant
(path coefficient: 0.002; t-statistic: –1.016). The non-significant effect
of structural formalization on creativity in the alliance team implies that
a rigid structure governing the alliance team is not important when
examining which structural factors promote or inhibit creativity within
the alliance team.
Hypothesis 4 suggested that greater structural formalization in the
alliance team leads to a low level of learning in that team. Supporting
this hypothesis, the empirical results show a negative and significant
effect of structural formalization on alliance learning (path coefficient:
– 0.169; t-statistic: –2.352). This result is consistent with the literature
which outlines that organizational structures that are characterized by
a high level of formalization impose boundaries and limit interaction
among members of a team; hence, limiting the learning that can occur.
DISCUSSION OF RESULTS
Results from the model estimation were consistent with the literature
with the exception of the result for Hypothesis 3—the effect of formalization on creativity within the alliance team. Implications of our empirical findings are that formalisation and centralisation have noticeably
different effects on the alliance team’s creativity. For example, introducing procedures to implement formalization has an effect that is different to that resulting from limiting decision-making (centralisation).
Alliance teams require a structure that reflects a low level of formalization, while employing creativity methods. However, decision-making
delegation within the alliance team is necessary in order for learning to
occur and thus is necessary if an alliance is to have dynamic capabilities. This suggests that the debate should not be focused on an either/
or application of centralisation or formality for the context of the
alliance team, but on an approach that combines the two aspects of
These findings are largely aligned with the earlier work of Burns and
Stalker (1961). Their work implies that mechanistic structures exhibiting high levels of centralization and formalization are those that are
most suited for stable environments where the role of dynamic capabilities is negligible. While their focus of analysis has been the firm, our
findings suggest that the logic underlying their framework applies also
to interorganizational structures that can be found in alliance teams.
Thus, organizational structures in alliance teams that can be described
as organic structures—characterized by low levels of formalization and
centralization and displaying characteristics such as loose organization, lateral communication, minimal control for task definition and
implementation, and delegation of decision-making to members of the
alliance team—would fit best dynamic environments requiring change.
Our findings support this notion and show that alliance teams that can
be characterized by an organic structure exhibit dynamic capabilities
evidenced in the underlying creativity and learning processes.
Examining the effects of those control variables included in our estimation (see Appendix 2) suggests that the creativity that occurs
within the alliance team is positively influenced by other factors than
organizational structure. For instance, intrinsic motivation of the
alliance team members, their engagement in critical thinking, the diversity of the team members, their attitude towards risk taking and the
communicative interaction within the alliance team all affect creativity.
In addition, creativity displayed in the alliance team is negatively affected by extrinsic motivation. These findings demonstrate that the organizational structure of the alliance matters but that managers need to
address additional aspects in order to strengthen creativity in the
alliance team and, ultimately, to enhance dynamic capabilities in the
Furthermore, the learning displayed within the alliance team is also
influenced by additional aspects characterizing the alliance team. Creativity exhibited in the team, intrinsic and extrinsic motivation of the
team members, critical thinking and communicative interaction
amongst the team members affect learning in the alliance. Hence, in
addition to the organizational structure depicting the alliance team,
managers can address other factors to enhance the learning in the
alliance team to improve the alliance’s dynamic capabilities.
Dynamic capabilities in the alliance are fuelled by a group of members
who are motivated, embrace diversity, engage in critical thinking and
communicate effectively. High levels of formalization and centralized
decision-making within the alliance team impede the development of
dynamic capabilities. These features of the alliance team that influence
the development of dynamic capabilities reflect those relating to organic systems (Burns and Stalker, 1961). Thus, our empirical findings support the theoretical assertions that organic structures are more suitable
for developing dynamic capabilities than mechanistic structures are.
Moreover, we demonstrate that this logic applies as well to the alliance
This study provides explicit insights into the relationships between
structural formalization and centralization and dynamic capabilities in
alliances. Theoretically, we have specified how 1/the literature on organizational structure and Burns and Stalker’s work (1961), in particular,
can be linked to the literature on dynamic capabilities, and 2/the link
between organizational structure and dynamic capabilities can be
examined within an alliance context. Both of these contributions fill an
important gap in the literature and provide a basis for additional studies on alliances and dynamic capabilities. For instance, future research
can examine the effects of organizational structure on dynamic capabilities in alliances conditional on other factors such as the stability and
durability of the alliance.
From a managerial perspective, this study provides important and
actionable insights as well. Alliance managers can further increase
creativity by reducing the concentration of decision making from a
small group of people to more widely dispersed members of the
alliance team. It is also important for managers to note that changing
the degree of formalization—the use of formal norms and procedures—within the context of the alliance team will not have a consequence, when considering creativity. This indicates managers are
more likely to accomplish better outcomes from the alliance team by
focusing on other areas rather than the levels of formalization present
when following procedures and reliance on formal policies.
The findings suggest that structural formalization and centralization
negatively influence alliance learning. This implies that managers
should focus their effort so that the alliance is organized loosely, people within the alliance team do not rely on formal policies in conducting their work, and that the alliance team is not very hierarchically
organized. Furthermore, it is important that members of the alliance
team are involved in determining objectives and have wide latitude in
making job-related choices.
In conclusion, organizational structure matters in alliances and affects
the development of their dynamic capabilities. Our findings provide
support for the arguments put forward by authors such as Burns and
Stalker (1961) who state that structure influences an organization’s
innovation processes such as learning and creativity. Our results suggest that mechanistic structures in alliance teams hinder the development of dynamic capabilities, whereas organic structures are more
conducive in these interorganizational settings.